Abstract

Info

Title:: Iris segmentation algorithm based on contour matching and polynomial fitting

Author(s):: ZHANG Wei1; LI Jing1; L?Xiaoqi1; 2; 1. School of Information Engineering, Inner Mongolia University of Science and Technology, Baotou 014010, China 2. School of Information Engineering, Inner Mongolia University of Technology, Hohhot 010051, China

Keywords:: Keywords: iris image segmentation contour matching polynomial fitting

Abstract:: Abstract: In the process of iris image acquisition, there are noises such as occlusion and declination due to individual differences, acquisition environment and other factors. Aiming at this problem, an iris segmentation method based on contour matching and polynomial fitting is proposed. First, the iris images are divided into ideal iris images and non-ideal iris images. Then, for ideal iris images, the iris segmentation is realized by detecting the iris boundary according to the standard equation of circle, and for non-ideal iris images, after obtaining the irises with occluded part through the contour matching algorithm, the polynomial fitting method is used to remove the occluded part such as eyelid to realize the accurate segmentation of irises. In this paper, the error rate of this iris segmentation is 1.5% for nice1 and 1.9% for nice2, and the F1 value reaches 93.65%. The results show that this method can effectively remove the occlusion noise in iris images and has a high accuracy of iris segmentation.